Cephalic phase of insulin secretion is regulated by autonomic and endocrine responses to food-related sensory stimulation such as sight, smell, and taste. Human taste perception comprises of at least five distinct qualities: bitterness, saltiness, sourness, sweetness, and umami, the sensation elicited by glutamate, commonly found in protein (meat, fish, and legumes) and flavor enhancer such as monosodium glutamate (MSG). Both the sweet and umami taste stimuli had been shown to illicit cephalic-phase insulin release in rats. Oral sensory stimulation in human with modified sham feeding (MSF ? where food is smelled, chewed, but not swallowed) had been shown to enhance insulin release during the cephalic phase, lower postprandial glucose level, and improve glucose tolerance in healthy subjects. The loss of pre-absorptive insulin response has been shown to impair glucose tolerance. Furthermore, patients with type 2 diabetes and their first degree relatives had been shown to have impairment of sweet taste. [unreadable] [unreadable] Recently, we found that glucagon-like peptide-1 (GLP-1) and peptide YY (PYY) have been found in the taste cells located in the taste buds of mice. These new findings raise several interesting questions of whether strict tasting of food without ingestion may stimulate secretion of GLP-1 and PYY from the taste cells, whether their secretion is involved in the afferent input of the cranial nerves, and whether this secretion is impaired in obesity and in patients with pre-diabetes or type 2 diabetes. We also want to investigate whether different tastants, such as sweet versus umami, and different food contents such as percent fat versus carbohydrate compositions, would elicit different hormonal responses. [unreadable] [unreadable] The purpose of this study is to determine if GLP-1 and PYY are indeed involved in the cephalic-phase response. If so, are there differences in response among healthy lean individuals, healthy obese individuals, and individuals with pre-diabetes or type 2 diabetes. Participants receive different tastants as interventions at each of their seven study visits and frequent blood samplings are performed before and after each intervention. We are actively recruiting healthy lean subjects, healthy obese subjects, subjects with pre-diabetes and subjects with type 2 diabetes for this study.